RCA 9T246 linearity
 

I've got an RCA 9T246 (or so it says) that lost vertical sync, it could not get slow enough.
Removing the case seems to have permanently cured that.
A resistor might have been shorting out, I moved it a bit.

I decided to realign and readjust it. The alignment was difficult, and
would change when put back in the case, which is metal. I was able to get
a bottom cover off and adjust a few coils with it in the case. This gives
a quite nice picture.

But the sweep linearities are abysmal and hard to adjust. The outer
areas are squashed. Its almost as if the yoke was misplaced. I've
got it so all centered circles are circular, but a crosshatch shows the
squashing. Swapping 6SN7s changes nothing.

Does anybody have any ideas to fix this?

The RCA manual for that set does not correspond to reality. It appears that the one for the 8T241 matches mine.

The problem with the linearity may be due to the case or case hardware being magnetized. A degaussing coil may work. There are also magnetic tape erasers that may work.

The nonlinearity does not depend on whether or not it in or out of the case.

In this chassis, leaky capacitors are generally the culprit.

A photo or two of the display would help pinpoint the component.

And the only differences I can think of between the 8T241 and 9T246 is the electro magnet to fixed magnet focus ring and the 5V4 to 6W4 damper swap.

OK, a picture. Sun almost shining on me. Its symmetric, but squashed at the edges.


http://www.videokarma.org/attachment...1&d=1654210798

Apart from the Height control being a smidgen too far advanced causing the bottom stretch and cramp in the upper third, a smidgen too little horizontal drive and the linearity misadjustment which when adjusted would bring out the sides, that looks pretty good. Is that the problem with it?

Well, that and also trying to get it centered at the same time without making the
dip in the upper left much much worse, yes. And, other than that, Mrs. Lincoln, how did you like the show?

It is indeed a juggling act. Forgot to add that when you make the adjustments mentioned, you will also obviously have to recenter the picture with the focus ring which means a touch up of the focus and ion trap. The focus ring requires finangling to avoid the neck shadow.

The horizontal linearity and drive readjustment will mean reducing width. And lastly, when adjusting the height, the vertical linearity must be readjusted also. I love analog electronics!

The CTC5 convergence compared to later color sets is also plagued with control interaction. I believe this was somewhat minimized in later designs by including clamping diodes in the convergence circuitry.

What do you mean by "focus ring"? It has a DC powered focus coil, which thankfully is well within range.

Its centered by adjusting the tilt of the yoke.

I recall seeing a 9T246 with a focus permanent magnet. It was mounted the same as the coil. There was no focus rheostats. Only a magnetic shutter to adjust focus. Centering was the same: the three sprung screws to tilt it.

Apologies for a quick jump off topic -

The diode clamping circuits in later sets keep the center DC convergence constant when the dynamic parabola waveform amplitudes are adjusted. (The sawtooth does't affect the center DC.) In the CTC-5, the points of constant DC adjustment when adjusting the parabolas are at the square root of 1/3 the distance from center to edge of raster (or about 0.577). This is why it is best to ignore the shift in center DC while adjusting the parabolas (and sawtooths) to make the center horizontal RGB lines parallel although not coinciding, and then re-adjust the static (DC) to get the whole line converged.

I threw in the CTC5 convergence as an analogy to the 9T246 because in a similar way having circle around between the DC static and dynamic adjustment interaction was like the linearity and positioning. But you made an important point as I recall CTC5 manual highlights the horizontal RGB lines parallel. I made the mistake of initially attempting convergence as one would with a later set. After learning to put off the static convergence until later, the work became easier and the results were much better.

In the same way in the 9T246, you aim for best linearity and width first then worry about centering, focus etc later.

I played with the vertical again yesterday and got it a bit better. Its still got problems at the bottom.
The vertical linearity control is all the way clockwise. The vertical was the circuit
that originally sick, but I never found anything wrong with it. I didn't measure every part out of the circuit
since it started working when removed from the case. I think the next step if to remove it from the
case (again :sigh: ) and check each and every cap and resistor, fix any that were even a tiny bit off, the power up
and use a scope.

I did the above tests and found that the fixed vertical output cathode resistor,
supposedly 2.7K, was 4k. I fixed that. It turns out that did not help since the
sweet spot is 4k total for that resistor and the pot.

I finally decided to see if it was the tube, a 6K6. I have no other 6K6 but I have
a couple of 6F6 ones, which have almost identical curves. There was no difference.

So I decided to try other tubes. First I tried a 6L6 (metal). This gave a perfectly
linear scope trace and with a touch of adjustment a very nice almost perfect linearity. I then tried a 6V6 an its almost but not quite as good.

The 6K6 is a .4 amp filament, the 6V6 .45, and the 6L6 .9 amp.
Since the currest is adjustment limited, the plate dissipation is about the same
in all the tubes and way under ratings. I'l probably leave the 6V6 in it.

But since its triode connected, why did they not use a triode?

In the late '40's there was no medium power triode. There was no need for one. Connecting a penthode as a triode lowered the plate impedance and it suited admirably.

Are you able to do a transconductance test on the 6K6? Or at least an emission test? I suspect the tube is bad.

And it would be nice if you could post the improved linearity.